Fidan Mahmudova
Baku
Summary
Robotics Engineer with hands-on experience in deep learning and computer vision, specializing in autonomous systems development. Strong background in both research (robotics navigation systems) and industry applications (automation and control systems).
Overview
3
3
years of professional experience
7
7
years of post-secondary education
Work history
RESEARCH ASSISTANT
Robotics Perception and Manipulation Lab
Minneapolis
2024.05 - 2025.08
Project/Thesis : "Object-Centric Local Navigation for Quadruped Robots"
Technical Contributions:
- Implemented vision encoder pipeline integrating ResNet18, DINOv2, and Grounded SAM2 (Grounding DINO + SAM2) with cross-attention mechanisms for RGB-based semantic object understanding.
- Developed encoder-MLP policy architecture for end-to-end learning from demonstration data, enabling centimeter-level positioning accuracy from RGB observations on Boston Dynamics SPOT robot.
Key Results:
- Achieved 98% success rate for object relocation tasks and 91% accuracy for cross-environment generalization, demonstrating robust vision-based navigation capabilities
GRADUATE RESEARCH ASSISTANT
Minneapolis Robotics Institute
Minneapolis
2023.08 - 2024.01
- Lead team for FTC at MN Stem Partnership
CONTROL & INSTRUMENTATION INTERN
BP
Baku
2022.07 - 2022.09
- Optimized control logic for riser valves in the Gas Export Pipeline
- Involved in Factory Acceptance Testing (FAT) and Site Acceptance Testing (SAT) for control systems
AUTOMATION INTERN
SOCAR Turkey
Baku
2022.06 - 2022.07
- Designed HMI screen using SIMATIC-S7 PLCSIM
- Tuned a PID controller for NHT process
Education
MS - ROBOTICS
University of Minnesota-Twin Cities
Minneapolis 2023.09 - 2025.08
B.ENG. - AUTOMATION ENGINEERING
Baku Higher Oil School
2018.09 - 2023.07
Skills
- Python
- PyTorch
- Deep Learning
- C
- Matlab
- Simulink
- Modern Control
- Robust Control
- ROS
- Raspberry Pi
Websites
Languages
Azerbaijani
Native
English
Proficient (C2)
Russian
Proficient (C2)
Accomplishments
- Team "Istiqlal" - 1st place winner in CanSat category
- Programmed Raspberry Pi to integrate and manage diverse sensors (GPS, IMU, etc.) for a Satellite model
Technofest Competition
Projects
OBJECT-CENTRIC LOCAL NAVIGATION FOR QUADRUPED ROBOTS | MS Thesis
Developed imitation learning pipeline with encoder-MLP architecture, implementing vision encoders (ResNet18, DINOv2, Grounded SAM2) with cross-attention mechanisms, achieving 98% navigation accuracy
LIDAR-GUIDED SAMPLE DELIVERY SYSTEM | CSCI5551
Implemented Bug 2 path planning algorithm for autonomous TurtleBot3 navigation across complex terrain using LIDAR perception
LUNG TUMOR SEGMENTATION | EE5561
Co-developed modified U-Net and Res U-Net architectures for accurate tumor detection in CT medical images
HUMAN-ROBOT INTERACTION WITH LLM | CSCI5541
Set up PyBullet simulation environment for empathetic social robot using large language models for context-aware responses
Timeline
RESEARCH ASSISTANT
Robotics Perception and Manipulation Lab
2024.05 - 2025.08
MS - ROBOTICS
University of Minnesota-Twin Cities
2023.09 - 2025.08
GRADUATE RESEARCH ASSISTANT
Minneapolis Robotics Institute
2023.08 - 2024.01
CONTROL & INSTRUMENTATION INTERN
BP
2022.07 - 2022.09
AUTOMATION INTERN
SOCAR Turkey
2022.06 - 2022.07
B.ENG. - AUTOMATION ENGINEERING
Baku Higher Oil School
2018.09 - 2023.07